Spatial Information Modeling For Indoor Map Matching And Evaluation Of Its Impact On Positioning Accuracy

Indoor positioning services are becoming more prevalent in our daily lives and at work.Researchers in related fields have done a lot of useful work to improve the accuracy of indoor positioning,among which map matching is an important method to effectively improve the accuracy of indoor positioning.Map matching is the process of matching the initial positioning data output by the positioning system with a map reflecting the surrounding physical environment,therefore improving the initial positioning accuracy.Map matching effectively applies the constraints of the indoor space layout to the initial positioning results,correct the initial positioning error.The researchers combined some methodologies with the concept of map matching to create a range of map matching methods that all improved the initial accuracy of indoor positioning.According to previous research,the quantity of current indoor maps is limited,and there are many mistakes that impact matching,making map matching approaches unsuitable for large-scale implementation.Before creating a map,it is crucial to first decide the content that the map should convey,as well as the level of detail that should be clearly expressed.However,the existing indoor spatial information model has not made corresponding regulations for the demand of map matching.The first issues to be considered in this research are what spatial elements should be included in an indoor spatial information model that may give significant support for map matching and how to express these elements.The researchers discovered that when the same map matching method was applied to different indoor scenes,the positioning accuracy of the same map matching method exhibited significant changes,and this difference had a significant impact on the map matching method’s implementation.At the same time,the evaluation of map matching methods is also extremely unfavorable.As a result,clarifying how the spatial characteristics made of indoor spatial elements impact the performance of the map matching approach is another important challenge for this work.So,how to get the indoor space elements mentioned above? Architectural engineering drawings are essential compilation materials utilized in the field of indoor mapping,however,there are still many issues with existing information extraction based on architectural engineering drawings,such as low efficiency and accuracy.Another important problem to be solved in this study is how to properly screen the massive data in the architectural engineering map to get the spatial elements necessary for map matching.The above problems all have an important impact on the implementation of map matching methods.If these problems cannot be effectively solved,the significance of map matching to indoor positioning will be greatly reduced.In response to the above problems,the key contributions of this work are expressed in the following three aspects:(1)For map matching,an indoor position-semantic model is constructed.Previous indoor spatial information models,when compared to the actual demands of the map matching method,were either too simple to provide effective support,or too complicated to operate,creating many hurdles to the map matching method’s implementation.In order to solve the above problems,this research constructs an indoor position-semantic model for map matching.This model is driven by the actual needs of map matching,and aims to express the position and semantics of spatial elements.The expression of topological relationships between spatial elements is presented.The expression of the indoor road network and grid directly required for map matching is derived,and the conversion between different coordinate systems is also taken into account.This information is important for map matching and have advantages over the other indoor spatial information models.(2)The influence rule of spatial characteristics composed of indoor space elements on map matching is clearly expressed.Different indoor environments provide different spatial constraints,so we often find that the performance of map matching methods in different indoor environments is quite different,which brings great trouble to our evaluation of map matching methods,but there are few studies on the influence of indoor space environment on map matching methods.This study starts from the mechanism of map matching method,analyzes in detail the influence of various indoor space characteristics composed of indoor elements on various mainstream map matching methods,and conducts effective verification.Compared with the vague expressions of traditional methods related to this research,the influence of spatial characteristics on map matching methods explored in this study is relatively clear,which provides a clearer basis for the objective evaluation and rational selection of map matching methods in related fields in the future,provides theoretical support for the mapping work of indoor map matching,and also has certain guiding significance for the improvement of indoor location-semantic models.(3)An automatic extraction method of indoor space elements for map matching is constructed.The traditional indoor automatic mapping methods are relatively inefficient,and cannot meet the needs of map matching in terms of quantity and accuracy.Aiming at the structural characteristics of architectural engineering drawings that place indoor space elements in layers,this research innovatively introduces the "Buffer" related method into the layer recognition work,which greatly reduces a large amount of redundant information.This is an advantage of this study over the previously indoor mapping studies.In the follow-up wall extraction work,in view of the complexity of the geometric shape of the wall structure,this study adopts the "Central Axis Transformation Algorithm" to extract the center line of the wall,which increases the robustness of this method and the accuracy of drawing.The automatic extraction method of spatial elements constructed in this study not only fully considers the needs of map matching methods in the implementation process,but also greatly improves the efficiency and accuracy of mapping compared with other traditional mapping methods,used for map matching.Therefore,our design method provides more solid guarantee and support the indoor map matching.To sum up,based on the needs of map matching,this study constructs an indoor position-semantic model specially oriented to map matching.Based on the previous study and associate challenges a new method for automatic extraction of indoor space elements for map matching is proposed.The research work in this study will provide a theoretical reference for the efficient implementation of map matching and the improvement of map matching methods in the future.Similarly,this study will also provide a relatively clear supporting basis for the objective evaluation of map matching methods.